Posted by Marvin from cache03.netway.at (195.96.10.36)on May 25, 1998 at 05:23:44:
In Reply to: Re: Update on BigRRed.... posted by Chris Lundie on May 24, 1998 at 15:48:59:
I don't really understand the question - did I
miss something? Redline offers vertex types with
iterated fog values (F) out of the box. So, to me,
the problem reduces to caculating the proper fog
values per vertex.... For table based fog this means
converting the w value of the vertex to an index
into the (stored) glide fog table, interpolating and
inverting the value and then putting it into
the vertex F.... The problem here is mostly the right
conversion from w to fog-index (problems: 1) is the
glide documentation right about the formula? 2) speed!!!)
I roughly (without interpolation yet) implemented this
concept when testing Excalibur, and I think it
basically works, however the w to index values
did not seem to be right in Excalibur (the w values
were always in about the same range) - I finally blamed
Excalibur for this...
In Barrage I got very thick fog and the whole
lighting was much too dark if I turned fog off....
Don't know how it should look like....
What I would need is either a 3dfx card
to see how these games should look like, or a game
that is D3D also, in order to compare the effect...
Turok declines to start, unfortunately - initially,
I planned to verify and adjust fogging routines
with the use of the Turok demo....!
Motorhead seems to use iterated alpha fog, not
fog tables, and I did not yet implement this
variant (is much easier!)....
The real problems would start, if a game uses the
special glide fog flags ADD2 and MULT2 for
multi-pass lighting calculation.....
But multi-pass rendering is a problem, anyway
(don't think Quake or something could be run
via the wrapper...) - at least I never dared to
think about that...
Since I still got no game with fogging really
right (did not try really hard yet, though, I still
wait for the right glide game to test it....),
please tell me: did I understand something
about glide/redline fogging basically wrong?
Or do I simply have to adjust the routines a bit
(my guess!)...????